establishing ongoing monitoring, surveillance and evaluation

11

ESTABLISHING ONGOING MONITORING,SURVEILLANCE AND EVALUATION

PROGRAMS

JEANNE MOLDENHAUER

Vectech Pharmaceutical Consultants, Inc.

Farmington Hills, MI

USA

INTRODUCTION

Once a system has been developed, appropriate procedures should be established tomaintain the system in a production environment. This includes establishingprocedures for routine and ongoing monitoring and surveillance, and evaluation of thedata obtained. This chapter provides an overview of these types of systems.

COMPONENTS OF THESE SYSTEMS

The environmental monitoring data that is generated should be collected in a mannerthat is in conformance with current Good Manufacturing Practices (cGMPs). Thepersonnel supervising the environmental monitoring program should be competent inthe appropriate scientific discipline and have the training (including within theindustry) and the authority required to make and implement change. Equipment that isused should be calibrated, systems should be appropriately validated and/or qualified,media should be properly prepared, and all operational procedures should be writtenand followed.

The established procedures should include appropriate controls to support theiruse. Cleaning of the facility and equipment, sanitization/disinfection agents andprocedures, sample site selection processes and specification of the frequency oftesting are key components to a good environmental monitoring program. Alert and

3www.pda.org/bookstore

4 EnvironmentaL Monitoring

Action levels should be established and based on individual sample sites, but one mayalso choose to specify the maximum number of microbial data deviations permitted inone area/system for a sampling period. The establishment of appropriate Alert and/orAction levels and a system for monitoring implies that all of the data obtained aresubject to continual review, and that Alert and Action level decisions are made bydesignated and authorized personnel qualified to make such decisions. To effectivelyexecute microbiological surveillance support systems require several componentsincluding:

• Documented system in place for identifying microbial data deviations

• Feedback mechanism for verification of any action taken in response to data

• All data should be documented

• Trending and/or data analysis

• Senior management should be involved, as appropriate

CLEANING AND SANITIZA TION/DISINFECTION

Implementation of cleaning and sanitization procedures is a critical component ofoverall facility control. Environmental monitoring data are used in routinelydetermining the ongoing effectiveness of these procedures. Care should be taken toensure that the cleaning personnel understand the criticality of the cleaning proceduresand know how to properly perform these tasks.

It is common for companies to have an internal auditor or supervisor periodicallyverify the cleaning procedures used, typically without the knowledge of the personnelperforming the tasks.

SAMPLE SITE SELECTION

Suitable sample sites vary widely depending on the cleanroom design andmanufacturing process. Careful evaluation of each process should be made in selectingsites. The primary purpose of sampling should be to provide meaningful, interpretabledata that can help identify actual or potential contamination problems associated withspecific procedures, equipment, materials, and processes. Following the validation and

www.pda.org/bookstore

Establishing Ongoing Monitoring, Surveillance and Evaluation Programs 5

qualification exercises, sites should be established for routine monitoring. Riskanalysis is recommended for performing this type of analysis. Some of the factors toconsider in selecting sites for routine surveillance are:

• Which sites can be identified where one would expect microbial contaminationmost likely to have an adverse effect on product quality?

• What sites would be most likely to demonstrate the heaviest microbialproliferation during actual production?

• What type of risk analysis procedure should be used to select the sites?

• What sites would represent the most inaccessible or difficult areas to clean,sanitize, or disinfect?

• What mechanisms are likely to disperse microbes in the environment?

• Would the act of sampling at a given site disturb the environment sufficiently tocause erroneous data to be collected or contaminate product?

• Should sampling only be performed at the end of the shift?

It may not always be practical to select a site at the most crucial identifiedlocation. One should consider whether crucial site monitoring would actually increaserisk of product contamination.

Other considerations in establishing an appropriate site for sampling include:

• Facility design

• Line configurations

• Validation data

• Manufacturing process information

• Historical data

• Test methodology, etc.

Selection of sites that are appropriate may change over time. On a long term basis,a company should establish procedures to re-evaluate the appropriateness of samplinglocations.


6 EnvironmentaL Monitoring

SAMPLING FREQUENCY

Frequency of monitoring requirements varies widely in the industry depending onseveral factors including, for example:

Type of manufacturing process

• Product

• Facility/process design

• Human intervention required

• Use of subsequent terminal sterilization

• Historical profiles of the microbiological environmental data

There is not one single sampling scheme that is appropriate for all environments.

Changes in sampling frequency, whether temporary or permanent, may berequired based on changes in practices, compendial requirements, development ofsignificant microbiological trends, and acquisition of new equipment, nearbyconstruction of rooms or utilities, and so forth.

Monitoring frequencies should be selected that can identify potential systemdeficiencies. The test frequency per site may be less frequent than the system or areafrequency (e.g., one may choose to rotate sample sites).

Prior to implementing any reduction in frequency, a summary of historical data,along with current and proposed sampling frequencies, should be reviewed andapproved by the appropriate Quality Assurance (QA) personnel. After preparing thesesummaries, data should be reviewed periodically to determine if the reduced samplingfrequency is still appropriate.

ALERT AND ACTION LEVELS

Most regulatory agencies require environmental monitoring programs haveestablished Action levels based on the applicable guidelines and the review ofhistorical data. Alert levels are also recommended. Some companies also choose to setlevels for individual cleanrooms or sample sites. Typically, the Action levels will bedriven by the regulatory or industry guidelines. The Alert levels will be driven byhistorical analysis of the environmental monitoring data. The application of Alert



and/or Action levels should be written and employed in a consistent, non-arbitrarymanner. To create consistency in treatment of Alert and/or Action levels, logicalinvestigatory and/or corrective action steps should be pre-specified. Records shouldshow that the microbial data deviation was recognized and that appropriate follow-upoccurred.

Once Alert and Action levels have been established, they should be periodicallyreviewed, as part of routine trend analysis. They may be revised to reflectimprovements, advances in technology, changes in use patterns, or other changes. Avariety of methods are used to establish these values.

DATA MANAGEMENT

Several different activities are part of data management, data collection, analysis ofdata, approaches used to analyze data, and interpretation of data. The routine reviewand analysis of environmental monitoring data is essential to aid in the interpretationof process stability and assess overall control performance. Management must be keptabreast of trends and the subsequent state of operations within their facilities.

Based on the large number of samples tested by a given facility, a computer-baseddata tracking system is recommended. Prior to implementation, all databaseapplications used should be validated/qualified for specific software applications.

DATA COLLECTION

Routine data must be pooled into a designated database or spreadsheet in a consistentrecord format. The record format should include:

• Monitoring date

• Specific sampling locations

Sampling methods

• Colony Forming Units (cfu) or number of organisms

• Non-viable count results

• Identification performed


8 Environmental Monitoring

• Product lot information

• Current Action! Alert level

The data tables can be completed using manual data entry, an image scannersystem, or electronic entry into handheld devices. Data integrity must be verified priorto analysis.

ANALYSIS OF DATA

Environmental monitoring trends are often difficult to obtain and recognize, due to thelow number of Colony Forming Unit (cfu) results usually obtained. Histograms,defined as pictorial graphs characterized by a number of data points that fall within acommon frequency, are one valuable tool. Different room classifications with definiterequirements will produce different histograms. Each area (or area type) andaccompanying data set must be viewed as distinct. A mathematical model should beapplied not only with the objective in mind, but also the type of data to be analyzed.

Grade A (ISO 5) areas where typical Action levels for samples may be one (1)microorganism, can be hard to trend. In these types of circumstances, rather than apure trend of data one may calculate a hit rate, i.e., how often do you get a countdivided by the number of samples taken. In this example, as the hit rate increases onewould assume an adverse trend. As the hit rate decreases the environment is 'cleaner'.

APPROACH TO DATA MANAGEMENT

When data is being analyzed, it is important to understand how the assessment will beconducted. For example, a generalized method that may be used for data to assess theenvironmental control is:

• Determine objective of analysis (e.g., analysis of site location data, examinationof a microbial data deviation, assessing the appropriateness of an establishedAction level, management update, and so forth).

• Specify the data set to be analyzed, e.g., one week, one month, only data from aspecific site.

• Apply data plots, such as histograms or pictorial plots, to access the basic data andto determine the nature of the distribution, if any. The data plots can also be usedto locate peculiarities such as outliers or patterns.



Observe the distribution and proceed with the appropriate mathematical modelthat best fits the overall objective. If data conform to a normal distribution, aparametric mathematical model may be applied. If the data is not consistent witha normal distribution, then a non-parametric approach may be applicable.

• In this example, an Action level at the 99th percentile is employed. Consistentwith the Action level at the 99th percentile are the following mathematical models(PDA, 2001). Models can only be applied if the data's character assumes adefinite distribution.

• Action level estimate for a data set reflecting an exponential or non normaldistribution = 4.6 x (mean cfu)

• Action level estimate for a data set reflecting a normal distribution = 2.33SD+ (mean cfu)

Note: When the Action Level is determined at the 99th percentile, an occasionalexcursion is expected due to the model applied. SD = Standard Deviation.

• Regardless of the statistical model chosen, the analytical method must beconsistent with the data and documented in the data summary along with results.

INTERPRETATION OF DATA

All data generated should be summarized and evaluated to determine whether theenvironmental monitoring process is in a state of control.

IDENTIFICA TION/CHARACTERIZA TION OF ISOLATES

Characterizing microorganisms recovered from environmental and personnelmonitoring is an important part of surveillance programs. The characterization systemselected by the laboratory should be defined in writing, including the frequency ofcharacterization and the standard procedures for the methods. Identification systemsthat are available have differing strengths and weaknesses. Procedures should carefullydescribe how the final identification is made when more than one identification systemis used (PDA, 2001).

Initially, many isolates may be characterized to establish baseline data of themicroorganisms found in the area.



Certain barriers apply to the characterization of environmental isolates. Themicroorganisms recovered from production environments may be highly stressed dueto physical factors such as limited nutrients, contact with chemicals or thermal stress.It may be difficult to obtain genus/species matches in identification system databases.The databases for commercial test kits and identification systems were designedoriginally for clinical isolates and may be incomplete with regard to industrial isolates;this may lead to misidentification of species or unidentifiable isolates. This area iscontinuing to be developed and enhanced (PDA, 2001).

It should be noted that for aseptic processes, Regulatory inspectors have placed astrong emphasis on the use of nucleic acid based identification systems, especially fortroubleshooting, root cause analysis, sterility test positives and media fill positives.

INVESTIGATIONS AND CORRECTIVE ACTIONS

When microbial data deviations occur, there can be a drift from the validated baseline.An investigation is needed to determine what happened and what needs to be done.Records should show that the excursions were recognized and appropriate follow-upoccurred.

The overall purpose of the investigative action is to establish, to the greatestdegree possible, a cause-effect relationship between the observed level ofenvironmental quality and causes for the excursions (i.e., sources of contamination).

To create consistency in the treatment of microbial data deviations, investigativeand/or corrective action, steps should be pre-specified in a written plan. A progressionof investigative!corrective actions or responses may be used in which sequential ormultiple excursions require greater consideration, than single or widely separateexcursions. Likewise, deviations which occur in areas that are critical to themanufacturing process may require a more rigorous investigation and corrective actionthan those which occur in areas that are judged less critical to the integrity of themanufacturing process.

DOCUMENTATION METHODS

A large amount of data must be collected for each sample site to ensure that the datacan be used in subsequent evaluations. Written procedures should identify the datarequirements, as well as what must be kept.



ENVIRONMENTAL MONITORING FOR TERMINALSTERILIZATION PROCESSES

An environmental control program designed for a terminal sterilization process shouldbe concerned with microbial flora that contributes to the spore bioburden of theproduct prior to sterilization. This includes testing of high purity water, distilled water,sterilizer cooling water, treated water and city water. Air, surfaces, and microbial levelsof containers and closures are also routinely monitored. While control of theenvironment in which the products are prepared is important, the most critical aspectof the program is the spore bioburden of the filled product to be sterilized. Appropriatecontrols for the manufacturing process ensures that the spore (heat resistant) bioburdenlevels present in the product sterilization cycle do not exceed the validated capabilitiesof the process and that the desired sterility assurance levels are achieved (PDA, 2001).

ENVIRONMENTAL MONITORING FOR ASEPTIC PROCESSES

The monitoring program for aseptic processes is specifically designed to determine thenumber and type of microorganisms associated with direct assembly, or preparation ofproduct, prior to sealing of the filled containers. The number of sample sites andfrequency of monitoring are generally greater than monitored for established terminalsterilization processes. Air, water, personnel, compressed gases, floors, walls,machinery, and other surfaces within the filling room are routinely monitored.Microbial levels are also routinely monitored on containers and closures. Adequateenvironmental control is an integral part of the aseptic manufacturing process.

ENVIRONMENTAL MONITORING AND ISOLATIONTECHNOLOGY

The environmental control program for aseptic filling isolator systems may be similarto that used for a conventional aseptic filling operation with the exception of surfaceand personnel monitoring. After sufficient data is collected, routine surface and airmonitoring may not be warranted if a validated sanitization cycle exists for the interiorsurfaces of the isolator. However, particulate air sampling would be performedroutinely. Surface monitoring may be used during initial validation runs to support theeffectiveness of the sanitization cycle and maintenance of clean isolator surfacesbetween sanitization cycles. If surface monitoring is performed it should be done afterthe completion of filling so as not to introduce any extraneous contamination orresidual growth media during the filling operation. Monitoring of personnel is notrequired for isolator systems; but monitoring of isolator gloves/half-suit is required.


12

WATER MONITORING

Environmental Monitoring

Water is the most widely used substance, raw material, or ingredient in the production,processing, and formulation of compendia1 articles. Control of the microbial quality ofwater is of great importance in the parenteral manufacturing facility since it may beused for formulating product, as well as for various washing and rinsing processes.Once a water system is validated to be in a state of control, appropriate samples shouldbe taken from the holding and distribution system to assess the microbiological qualityfor its intended use. In addition to requirements for pharmaceutical manufacturing,there are also requirements that must be met for drinking water in most countries.

The time elapsing between collection of water samples and examination typicallyshould not exceed 24 hours. Each site should have validation data to support the timelimitations imposed for collection and testing of samples (PDA, 2001).

MONITORING OF COMPRESSEDGASES

Compressed air or other gas that is used, especially in an aseptic environment, shouldbe tested as part of the normal room air on a frequency that would assure that the gasdoes not adversely effect the environment. Testing may be performed at a lowerfrequency than routine testing for monitoring the environment.

Nitrogen, which is heavier than air, is used as a blanket over some productformulations and in some tunnels. Argon, lighter than air, may also be used. However,the weight difference of the gas being sampled must be taken into consideration. Theflow device used will have to be reset or designed to account for any differences, orthere may be less or more gas in the sample than is recorded on the flow device.Accurate sampling of gases requires that one considers many parameters including:

• Elevation, i.e., any gas or air expelled into the environment should be at a levelsimilar to that of the room.

• The typical aseptic manufacturing environment incorporates either a validated0.45 or 0.2 /.lm filter. Testing may be performed as a confirmation of the filter'sacceptability, as a test of the filter's completeness and not a test of the compressedair/gas system itself. Some operations may not incorporate a filter at point of use.A pre-filter may be installed in the system to reduce the bioburden to the point ofuse filter that will extend its life, and reduce the possibility of highercontamination reaching the area.

• The sterilization of components prior to use in the test instrument is critical.


Establishing Ongoing Monitoring, SurveiLLanceand Evaluation Programs 13

• All compressed air connections, which do not affect the air to the workspace,should be monitored with less frequency; however, any connection whichintroduces air to the environment should be monitored on a frequency so as toassure the conditions of the aseptic environment.

Precautionary measures should be taken when sampling gases which may have anexplosive nature.

• A medium used for evaluation and incubation and rendering evaluations shouldfollow the standard practice as for routine monitoring sites (PDA, 2001).

AIR MONITORING

A comprehensive environmental monitoring program should include routinemonitoring of both viable and non-viable airborne particulates. Viable particulates aregenerally of significant concern in sterile product manufacturing environments;however, non-viable particulates should also be monitored as another source ofpotential product contamination. Current techniques for monitoring viable particulatesin air are limited by: the equipment available; the time necessary to demonstrate thepresence of microorganisms in the sample of air taken; the inability to re-sample theenvironment in a timely fashion when results warrant; and difficulties in continuouslymonitoring the environment due to considerations, such as drying out of the recoverymedia. With the advancement of rapid microbiological methods, these concerns maybe minimized.

Non-Viable Particulate Monitoring

Monitoring of non-viable airborne particulates is a necessary component of anenvironmental monitoring program. Such monitoring demonstrates control ofpotential contaminants in the environment to which the product, during themanufacturing process, is exposed. Classification of production areas is generallymade based upon the level of non-viable particulates in the air. Classified areas mustcontinually meet these established particulate levels.

Viable Monitoring

Microorganisms present in the air are generally associated with solid or liquidparticles. These particles may consist of a single unattached cell or, more commonly,as clumps of organisms. Organisms may adhere to a dust particle or other 'raft', or, ifunattached, exist as a free-floating particle suspended in the air. These particles may



remain suspended in the air for extended periods of time due to the local air currents,but vertical unidirectional airflow, as well as gravity, will move them towards the flooror the nearest return air grill. (This assumes low-mounted returns, which may notalways be the case. Fermentation areas may use high-mounted returns.)

SURFACEMONITORING

In addition to conducting viable air monitoring to determine the microbial bioburdensurrounding the manufacturing operations, surface monitoring is conducted todetermine the microbial bioburden of surfaces within the manufacturing area, as wellas on equipment and product contact surfaces.

PERSONNEL MONITORING

Personnel introduce contamination in an aseptic environment, due to themicroorganisms and non-living particulates present. It is essential that all employeesentering an aseptic environment be carefully selected and trained. Training shouldinclude personal hygiene, an introduction to microbiology, aseptic techniques, andmicrobiological gowning certification. After an individual has been trained andcertified, routine microbiological monitoring of garments and finger impressionsshould be completed to assess the ongoing practice of aseptic technique. All trainingand certification activities should be documented and kept as part of the employee file.

MONITORING PRODUCT OR COMPONENT BIOBURDEN

Bioburden testing is performed on non-sterile product to determine its microbial load.The intended use of the product will dictate the allowable acceptance levels and theexclusion of objectionable microorganisms at the time of formulation. Listed beloware some factors that may impact product or component bioburden:

Raw material source

• Water

• Components

• Manufacturing environment


Establishing Ongoing Monitoring, Surveillance and EvaLuation Programs 15

• Processing of the formulation product

• Equipment

• Antimicrobial activity

• Water activity

For many terminally sterilized products, bioburden counts alone do not providesufficient information. It may be necessary to assess the thermoresistance or D-valueof the bioburden. Total bioburden counts that are within limits may cause a significantproblem if the bioburden present exceeds the thermoresistance anticipated for thesterilization model.

D-values can be determined using sophisticated equipment (thermoresistometer)with square wave heating, and heat-up and cooling times less than or equal to 10seconds. For routine screening ofbioburden, a heat shock or boiling water test is usedto rule out the presence of organisms exceeding a predetermined D-value.

PARAMETRIC RELEASEAND BIOBURDEN

Parametric release is a term used to describe sterilization processes that are so tightlycontrolled and understood that it is possible to ensure sterility of the product when theparameters are met, even without performing a sterility test. The allowance ofpharmaceutical manufacturers to use parametric release under the approval of FDA in1985 increased the importance ofbioburden testing, characterization, and resistance ofrecovered microorganisms. FDA Compliance Policy Guide (7132a.13), issued in 1987,details the criteria that must be met for Parametric Release.

Major emphasis is placed on the thermal resistance of recovered spore formingmicroorganisms. Recovered spore formers with greater resistance than the indicatororganism used in the cycle validation, would render the batch to be deemed non sterile.

IN-PROCESS TESTING

In-process environmental monitoring samples are taken to evaluate variousparameters, including the ability of the equipment to perform within specifiedenvironmental quality standards; the operator's ability to maintain area cleanlinessduring process operations; and the effectiveness of cleaning for the facility and itsequipment. This monitoring is typically performed in production areas, and during



operations, where product is exposed to environmental or operator contamination;however, it is not always included for closed systems, as the results may not have acorrelation to product impact. Process-related monitoring may include surface and airsites near aseptic connections or product transfer steps. The manufacturing operationsmonitored may occur in an open room, under laminar flow, or within a 'closed'system. Sites should be chosen to demonstrate process integrity in both 'open' and'closed' processes. Sample sites and levels should additionally be chosen to providemeaningful data about a given operation. As an example, non-viable particle countstaken during loading of powdered media into a vessel in a Class 100,000/ISO S/GradeD area may not provide data that is indicative of process quality. Non-viable particlecounts taken during aseptic processing operations (excluding powders) in Class100/ISO S/Grade A, areas may provide more valuable information about processcontrol.

The subsequent purification/bioburden reduction steps in a process may alsoimpact the degree to which in-processing testing is warranted. Test frequencies forbatch-related, in-process monitoring may differ from those for routine areamonitoring. In many cases environmental monitoring, performed in conjunction withbatch production activities, may fulfill the requirements for routine area monitoring.

Surface and viable air samples that select for the host organism may beappropriate in a fermentation/recovery process area. This data may help todemonstrate process integrity and/or cleaning effectiveness during a productchangeover procedure.

ENVIRONMENTAL MONITORING DURING ROUTINESTERILITY TESTING

Sterility testing facilities should be designed and operated in an equivalent manner toaseptic processing areas. Environmental monitoring should be conducted in an activemode during each shift, with Alert and Action levels set that are comparable to thoseused in aseptic process areas in the manufacturing plant. Monitoring should beconducted to demonstrate continuous microbial contamination control, consistenttechnician performance and to obtain information concerning the possible source ofthe microorganisms associated with sterility failures.

REFERENCES

PDA (September/October, 2001) Technical Report B Revised-Fundamentals of anEnvironmental Monitoring Program, PDA J. Sci and Techno/. 55: 5.


establishing ongoing monitoring, surveillance and evaluation

Documents